There is an estimation error caused by the non-independence of the source signal in the independent component analysis (ICA) method. Thus a system harmonic impedance estimation method based on the hybrid optimization and the separation consistency judgment is proposed. First, the data of the harmonic voltage and harmonic current measured at the point of common connection are centralized and whitened. Secondly, the position of the initial solution is obtained by chaotic mapping, and then the separation matrix is obtained by the biogeography-enhanced fireworks hybrid algorithm with the objective function of maximum negative entropy. Further, the separation matrix with better independence is obtained through separation consistency judgment. Finally, the harmonic impedance of the system is calculated by the least squares method. Simulation and case study results prove that the proposed method has better estimation accuracy than exiting ICA algorithms. This work is supported by the National Natural Science Foundation of China (No. 51977127).